How Is Ozempic Made? From Yeast Cells to Injection Pen

Ozempic starts its life inside genetically engineered baker’s yeast, the same species used to brew beer and leaven bread. Novo Nordisk inserts synthetic DNA into Saccharomyces cerevisiae so the yeast produces a precursor version of semaglutide, the active drug. That precursor then goes through a long chain of purification steps, chemical modifications, and quality checks before it ends up as the clear liquid inside the familiar prefilled injection pen. The journey from a single yeast colony to a finished pharmaceutical product is more intricate than most patients realize, and every step exists to solve a specific biological or chemical problem.

Why Yeast, and How the Cells Are Programmed

Semaglutide is a peptide, a short chain of amino acids that mimics a natural human hormone called GLP-1. Your body makes GLP-1 in the gut after eating, and it signals the pancreas to release insulin while also slowing digestion and reducing appetite. The trouble is that natural GLP-1 breaks down in minutes, far too quickly to be useful as a drug. So semaglutide is a modified version, sharing about 94% of its structure with human GLP-1 but engineered with specific amino acid swaps and a fatty acid attachment that let it survive in the bloodstream for about a week.¹

To manufacture this peptide at scale, Novo Nordisk uses recombinant DNA technology. Scientists design a small loop of DNA called a plasmid that carries the genetic instructions for producing a semaglutide precursor. This plasmid is inserted into S. cerevisiae yeast cells. The plasmid encodes a fusion protein: it includes a signal sequence (called the mating factor alpha pre-pro leader) that tells the yeast’s cellular machinery to secrete the protein outside the cell, along with a built-in cleavage site. As the fusion protein travels through the yeast cell’s internal processing compartments, an enzyme called Kex2 endopeptidase cuts it at exactly the right spot, releasing the mature GLP-1 precursor peptide into the surrounding liquid.²

This secretion trick is one of the key reasons yeast was chosen over bacteria like E. coli. When the precursor is released directly into the fermentation broth rather than locked inside the cell, the downstream cleanup is considerably simpler. You skip the step of breaking cells open and dealing with the tangled mess of intracellular proteins. The yeast essentially does part of the purification work for you.

Growing the Yeast at Scale

Once the engineered yeast strain is ready, it is transferred to a fermentor, a large stainless-steel vessel designed for precise control of temperature, oxygen, pH, and nutrient delivery. In a production-scale process model, the main fermentation takes place in a vessel of about 2,500 liters. The process begins with a batch phase lasting roughly 12 hours, during which the yeast grows on glucose at 30°C with air supplied at a controlled rate.³

After the initial glucose runs out, the process switches to fed-batch mode. Fresh medium is dripped into the vessel slowly, keeping the yeast fed but not overfed. Overfeeding yeast causes it to ferment glucose into ethanol rather than using it for growth and protein production, which would waste resources and potentially damage the peptide. In the process model, this fed-batch phase runs for about 60 hours with medium added at roughly 1.3 liters per hour, all while temperature and aeration stay constant.³

Research-scale studies use a similar strategy but with smaller bioreactors. In continuous cultivation experiments for GLP-1-like peptides, researchers have maintained yeast cultures in chemostat mode at a dilution rate of 0.08 per hour, with the pH held between about 5.3 and 6.5 depending on the strain lineage, and stirring speeds between 1,000 and 1,300 rpm.⁴ These details matter because even small changes in pH or feeding rate can alter how much precursor the yeast secretes and how cleanly it folds.

Harvesting and Purifying the Precursor

When fermentation is complete, the broth contains the semaglutide precursor peptide mixed with yeast cells, leftover nutrients, metabolic byproducts, and host-cell proteins. The first job is to separate the peptide from all that biological clutter. The broth is typically clarified by centrifugation or filtration to remove the yeast cells, leaving a liquid rich in the secreted precursor.

From there, chromatography is the workhorse. Think of chromatography as a molecular sieve: the liquid is passed through columns packed with materials that grab onto the peptide of interest while letting impurities flow through, or vice versa. Multiple rounds of chromatography using different separation principles are standard for pharmaceutical peptides. The goal is to reach the extreme purity levels that injectable drugs require.

Researchers are actively working on alternatives to reduce cost and complexity. One approach uses macroporous resins rather than traditional chromatographic media to capture the semaglutide precursor peptide. A study evaluating several resins found that one called XAD7HP achieved a maximum adsorption capacity of about 164 milligrams of precursor per gram of resin at 25°C, with a recovery rate of 75% and a desorption efficiency above 96% using an ethanol solution. The resin held up well through 15 cycles of adsorption and desorption, suggesting it could be a practical and reusable tool for large-scale purification.⁵

Chemical Modifications That Make It Last

The precursor peptide that comes out of the yeast is not yet semaglutide. It needs two critical modifications to become the long-acting drug that patients inject once a week.

The first modification happens at the amino acid level. Semaglutide has 31 amino acids, and two positions differ from natural GLP-1. At position 8, the natural amino acid alanine is replaced with a non-natural one called alpha-aminoisobutyric acid (Aib). This swap shields the peptide from an enzyme called DPP-4, which normally chews up GLP-1 within minutes. At position 34, lysine is replaced with arginine.¹ Some of these substitutions may be encoded directly in the DNA inserted into the yeast, while others are introduced chemically after the precursor is purified.

The second and arguably more important modification is the attachment of a fatty acid side chain to lysine at position 26. This C-18 fatty acid chain acts like an anchor: once injected, it grabs onto albumin, the most abundant protein in your blood. Albumin is large, circulates slowly, and is not easily cleared by the kidneys. By hitching a ride on albumin, semaglutide avoids rapid elimination and achieves a half-life of roughly a week, which is what makes once-weekly dosing possible. This albumin-binding trick is the key pharmacological innovation that separates semaglutide from the older GLP-1 drugs that required daily shots.

Attaching the fatty acid side chain is a delicate piece of synthetic chemistry. It must bond to the correct lysine and nowhere else, and the linker connecting the fatty acid to the peptide must be precise. This step is performed after the biological production phase, blending biotechnology with traditional medicinal chemistry.

Recombinant Versus Chemical Synthesis

Not every manufacturer uses yeast. Semaglutide can also be built entirely through solid-phase peptide synthesis, a chemical approach where amino acids are added one at a time to a growing chain attached to a solid bead. This method is flexible and does not require living organisms, but it generates significant chemical waste, needs extensive purification, and becomes less efficient as the peptide chain gets longer.

The recombinant approach (using engineered cells) has its own complications but offers higher yields for longer peptides. Researchers have reported yields of roughly 3 grams per liter of the semaglutide precursor using yeast-based gene recombination. A newer strategy using E. coli with a specialized chemical cleavage system pushed that even further, achieving about 5.85 grams per liter of the semaglutide core peptide while avoiding some of the refolding headaches that bacterial production normally entails.⁶

Traditional peptide synthesis also raises environmental concerns. The solvents, coupling reagents, and protecting groups used in solid-phase synthesis produce substantial waste per gram of product. As demand for GLP-1 drugs has surged, the push toward greener and more sustainable manufacturing methods has intensified.⁷ Recombinant production in yeast or bacteria, while not waste-free, typically consumes fewer harsh chemicals per dose.

Formulation and the Injection Pen

Once the semaglutide molecule is fully assembled and purified to pharmaceutical grade, it must be formulated into a stable, injectable solution. The drug is dissolved in a buffered liquid, usually containing a phosphate buffer and a preservative, at a carefully controlled pH. The choice of buffer and pH is not arbitrary: studies have shown that pH is a key factor in semaglutide’s thermal stability, and different pH conditions produce different degradation impurity profiles.⁸

The formulated solution is filled into glass cartridges, which are then loaded into the prefilled injection pens that patients receive. The Ozempic pen is a multidose device: each pen contains enough semaglutide for several weekly injections, with a dial mechanism that lets patients select the prescribed dose. The pen uses a fine needle that attaches before each injection and is discarded afterward. Human factors engineering plays a real role in pen design. During development of similar autoinjector devices, researchers found that use errors dropped significantly when instruction materials were refined and the device design was updated based on how real participants handled it.⁹

Keeping It Stable From Factory to Fridge

Semaglutide is a peptide, and peptides are finicky about temperature. The molecule retains its functional three-dimensional shape, an alpha-helical structure, up to about 60°C. Beyond that, things unravel quickly: at 80°C, the alpha-helical content drops to essentially zero, meaning the drug would lose its ability to bind to GLP-1 receptors and do its job.¹⁰ Even at more moderate elevated temperatures, degradation products accumulate over time.¹¹

This is why Ozempic pens are shipped and stored under refrigeration, typically between 2°C and 8°C. After first use, a pen can be kept at room temperature (up to about 30°C) for a limited period, usually up to six weeks, according to the product labeling. The cold chain, the unbroken line of temperature-controlled transport from factory to pharmacy, is a critical and expensive part of bringing any biologic drug to market. Breaks in the cold chain can degrade the peptide and generate impurities, some of which could trigger unwanted immune responses.

Quality Control and the Problem of Impurities

Pharmaceutical-grade semaglutide must meet extremely tight purity specifications. But no manufacturing process is perfect, and even small variations in production can introduce impurities. These are not just an academic concern: trace metals, for instance, have been shown to promote the formation of high-molecular-weight protein aggregates in both semaglutide and liraglutide. Some of these new impurities could theoretically trigger immune responses, as in silico analysis (computer-based prediction of how the immune system might react) has flagged certain degradation products as potential T cell epitopes. In Novo Nordisk’s own clinical trials, antibody responses to semaglutide were evaluated using a tiered analysis system to monitor for immunogenicity.¹²

The impurity question becomes even more pointed when follow-on (essentially generic) versions of semaglutide enter the picture. Comparisons between originator and follow-on semaglutide products have found that follow-on versions sometimes contain new impurity patterns, including different trace metals, residual solvents, and high-molecular-weight proteins not seen in the originator. Some of these novel impurities contained sequences flagged as potential immune triggers. Follow-on liraglutide products also showed a greater tendency to form fibrils, thread-like protein aggregates associated with reduced stability.¹³ This is a reminder that “same molecule” does not automatically mean “same product” when it comes to biologically produced drugs. The manufacturing process itself is part of what defines the product.

Scaling Up to Meet Global Demand

One of the most visible real-world bottlenecks for Ozempic has been supply. The explosion of demand for GLP-1 drugs, driven by both diabetes treatment and weight management, has outpaced manufacturing capacity. Producing peptide therapeutics at the scale needed to reach hundreds of millions of patients worldwide requires enormous investment in specialized production infrastructure.¹⁴

Novo Nordisk has been building new factories and expanding existing ones in Denmark and the United States, but biological manufacturing facilities take years to design, construct, validate, and bring online. Every fermentor, every chromatography suite, every fill-and-finish line must meet pharmaceutical regulatory standards before a single dose can ship. The company has invested billions of dollars in expansion, and competitors developing their own GLP-1 drugs face the same infrastructure challenge. This is not a problem that can be solved by simply ordering more raw materials; the bottleneck is the highly specialized facilities and the trained workforce to run them.

The Oral Semaglutide Twist

Injectable Ozempic is not the only form of semaglutide on the market. Novo Nordisk also sells an oral version called Rybelsus, which uses a completely different delivery strategy to get the same peptide into the bloodstream through the stomach. The tablet contains semaglutide co-formulated with an absorption enhancer called SNAC (sodium N-[8-(2-hydroxybenzoyl) amino] caprylate). SNAC raises the local pH in the stomach near the tablet, which protects the peptide from gastric enzymes and promotes its absorption across the stomach lining.¹⁵

The semaglutide molecule itself is manufactured the same way for both the injectable and oral forms. The difference is in the final formulation step: instead of going into a liquid-filled pen, the semaglutide destined for Rybelsus is combined with SNAC and pressed into a tablet. Oral bioavailability of peptides is notoriously low, so oral semaglutide requires a much higher dose in the tablet than what ends up in the bloodstream. Patients also need to take it on an empty stomach with minimal water and wait before eating, because food and excess fluid interfere with the SNAC absorption mechanism. The manufacturing challenge for oral semaglutide is less about making the peptide and more about ensuring consistent absorption from one dose to the next.

Why Manufacturing Defines the Drug

For small-molecule drugs like aspirin or metformin, the manufacturing process is relatively straightforward and the final product can be fully characterized by its chemical structure. If two factories produce the same small molecule, the products are functionally identical. Biologic drugs like semaglutide are different. The three-dimensional folding of the peptide, the purity profile, the types and amounts of trace impurities, and even the aggregation tendencies are all shaped by the specific conditions under which the drug was produced. A subtle change in fermentation pH, purification resin, or storage temperature can alter the impurity landscape in ways that matter clinically. This is the fundamental reason that regulatory agencies treat biosimilars differently from generic small-molecule drugs, requiring more extensive comparative testing before approval. For the patient uncapping an Ozempic pen, the manufacturing process is invisible. But every parameter in every step, from the DNA sequence inserted into the yeast to the buffer pH in the final vial, is part of what makes the drug work safely.